Use VCS-driven workflow
In addition to the CLI-driven workflow, HCP Terraform offers a VCS-driven workflow that automatically triggers runs based on changes to your VCS repositories. The CLI-driven workflow allows you to quickly iterate on your configuration and work locally, while the VCS-driven workflow enables collaboration within teams by establishing your shared repositories as the source of truth for infrastructure configuration.
In this tutorial, you will configure a VCS integration for your organization, connect your workspace to a VCS repository, and trigger a speculative plan based on a pull request. Then, you will merge the pull request to automatically apply changes to your infrastructure using HCP Terraform.
Configure new GitHub repository
In order to trigger HCP Terraform runs from changes to VCS, you first need to create a new repository in your personal GitHub account.
In the GitHub UI, create a new repository.
Name the repository learn-terraform
, then leave the rest of the options blank and click Create repository.
Copy the remote endpoint URL for your new repository.
In the directory of your cloned configuration, update the remote endpoint URL for your repository to the one you just copied. This updates your configuration to use your personal repository rather than the HashiCorp one you cloned the configuration from.
$ git remote set-url origin YOUR_REMOTE
Next, open your terraform.tf
file and comment out the cloud
configuration block.
terraform.tf
terraform {
/*
cloud {
organization = "organization-name"
workspaces {
name = "learn-terraform"
}
}
*/
required_providers {
aws = {
source = "hashicorp/aws"
version = "~> 3.28.0"
}
}
required_version = ">= 0.14.0"
}
When using the VCS-driven workflow for HCP Terraform, you do not need to
define the cloud
block in your configuration.
Now, add your change.
$ git add .
Commit the change.
$ git commit -m "Remove cloud block"
Finally, push your configuration to your personal repository.
$ git push
Enable VCS integration
To connect your workspace with your new GitHub repository, go to the workspace's Settings, then select Version Control.
On the next page, click Connect to version control.
Next, choose the Version Control Workflow option.
Select Github then GitHub.com (Custom) from the dropdown menu.
A new window will open asking you to authorize HCP Terraform to use your GitHub account. Click the green Authorize button to authorize the connection.
You may be asked to install HCP Terraform for your GitHub account or organization. If prompted, select your account or organization.
Next, search for and select your new learn-terraform
repository.
On the next page, scroll to the bottom, then confirm that HCP Terraform enabled Automatic speculative plans, then click Update VCS settings.
Speculative plans are non-destructive, plan-only runs that show you the changes Terraform will make to your infrastructure if you merge a pull request. The runs will not appear in your HCP Terraform logs and you can only access them through a direct link, which HCP Terraform will attach to your pull request.
Any pushes to the main branch of your repository will trigger Terraform runs in your workspace. You also enabled speculative plans, which will preview the changes Terraform will make to your infrastructure if you merge a pull request. HCP Terraform also lets you configure your workspace to trigger runs on changes to specific paths within your repository, or whenever you push a tag of a specified format.
Discard run
When you connected your workspace to your new repository, it triggered a new
HCP Terraform run to revert your instance type to t2.micro
, as defined by
your workspace's instance_type
input variable. Discard the run to avoid reverting the change.
Navigate to your workspace's Variables page and delete the instance_type
input variable by clicking on the ellipses next to it, then selecting
Delete.
.
Update configuration
In your repository directory, create a new file named terraform.auto.tfvars
and paste in the following contents.
terraform.auto.tfvars
instance_type = "t2.micro"
In addition to command line flags, variable sets, and workspace variables, you
can pass variables to your configuration in HCP Terraform through files named
*.auto.tfvars
. The variable value defined here will attempt to update your
instance type back to t2.micro
. Note that auto.tfvars
files have the lowest
precedence in a workspace, and any workspace specific variables will override
the values defined in this file.
Check out a new branch for your change.
$ git checkout -b use_auto_vars_file
Add your changes.
$ git add .
Commit your change.
$ git commit -m "Add auto.tfvars file"
Finally, push up your change to your new branch.
$ git push -u origin use_auto_vars_file
Enumerating objects: 162, done.
Counting objects: 100% (162/162), done.
Delta compression using up to 16 threads
Compressing objects: 100% (86/86), done.
Writing objects: 100% (162/162), 32.39 KiB | 32.39 MiB/s, done.
Total 162 (delta 70), reused 152 (delta 67)
remote: Resolving deltas: 100% (70/70), done.
remote:
remote: Create a pull request for 'use_auto_vars_file' on GitHub by visiting:
remote: https://github.com/redacted/learn-terraform/pull/new/use_auto_vars_file
remote:
To https://github.com/redacted/learn-terraform
* [new branch] use_auto_vars_file -> use_auto_vars_file
Branch 'use_auto_vars_file' set up to track remote branch 'use_auto_vars_file' from 'origin'.
Create a pull request
Visit your learn-terraform
repository on GitHub and open a new pull
request for your branch.
After your create the pull request, HCP Terraform will trigger a speculative plan. View this plan by clicking on the Details link in the Checks portion of your pull request.
When the speculative plan finishes, you and your teammates can review it before approving or merging the pull request. This HCP Terraform workflow enables teams to collaborate on planned infrastructure changes.
You cannot apply speculative plans, since your infrastructure would differ from the configuration on your main branch of your connected repository. You must merge the pull request to apply this change.
Merge pull request
Return to the GitHub UI, and merge the pull request with the Merge pull request button.
Switch back to the Runs page for your workspace in HCP Terraform. HCP Terraform has picked up the change to your configuration and started a new run.
Review the run details. Confirm & Apply the run. Once it completes, you can verify that Terraform updated your instance configuration by searching for the EC2 instance in the AWS console.
Next steps
The Terraform documentation contains more details on state, the VCS-driven workflow, and speculative plans.
In these tutorials, you configured a new workspace, provisioned infrastructure, and updated that infrastructure. Now, you are ready to destroy the resources and workspace.
In the next tutorial, you will clean up the resources created by your configuration and find additional learning materials on Terraform's configuration language.